COLOR PHOTOGRAPHY. 187 



appearing on the front surface. In transmitted light they appear in 

 part even more distinctly. 



This is a second proof that the colors are body colors, that is to say, 

 produced by absorption. 



This observation has been repeatedly made by other investigators, 

 but I have never yet found the conclusion drawn from it with regard to 

 the nature of the colors. Perhaps this is due to a fundamental error 

 which Zenker, the founder of the interference theory of color photog- 

 raphy, made in this connection. In his treatise on photochromie he 

 says, on page 81, in reference to the color reproductions through the 

 formation of stationary light waves : 



Similarly, it is natural that the same colors should appear by trans- 

 mitted light that are observed by reflection, for since the transmitted 

 light is certainly not the direct continuation of the incideut ray, but at 

 least in part also experiences several reflections, those same colors must 

 preponderate in it that correspond to the distances apart of the point 

 layers, that is, colors identical with those in the ray ordinarily reflected. 



By point layers are meant the elementary mirrors which are formed 

 in the sensitive film by the action of the stationary light waves. 



The colors, however, which are due to reflections from the elemen- 

 tary mirrors must be complementary to the reflected colors at the same 

 parts of the plate as in all pure interference colors, for they must 

 together make up the incident white light. They can indeed not fail 

 of this, for by hypothesis they are produced wholly by interference and 

 not by absorption. 



If one inquires how in the same difference of path, that is the double 

 distance between two neighboring elementary mirrors, different inter- 

 ference colors can be produced in reflected and transmitted light, he 

 forgets the phase changes occurring in reflection. At the same geomet- 

 rical plane where a ray of reflected light at the first elementary mirror 

 is thrown back in passing into optically denser or rarer parts, respec- 

 tively, the transmitted and twice-reflected ray is thrown back in pass- 

 ing into optically rarer or denser parts, respectively, and receives, 

 therefore, an opposite phase change. That at the second mirror is in 

 both cases similar. Thus there remains a phase difference of a half- 

 wave length, which causes the complementary coloring of the trans- 

 mitted light. There is no change in this relation with a greater number 

 of reflections. It may be objected that the phase change on reflection 

 at an elementary mirror must be the same, whichever side the light falls 

 upon it. That is the case; but it must be remembered that the ele- 

 mentary mirror is not a geometrical plane, but a layer of finite thick- 

 ness. Otherwise it could not, in the absence of absorption, reflect 

 light. 



Exactly this objection aids in determining the phase change on 

 reflection at an elementary mirror and not at a geometrical plane 

 coincident with its boundary or within, as was discussed above. Since 

 in transmitted light the twice-reflected ray experiences, with reference 



